Why is the past easier to predict than the future?
Retrospective Coherence is looking back in time and creating a coherent narrative around what happened. The accuracy of this narrative is suspect because the narrative can only be based on assumptions of the motivations and methods of decision making that happened in places where no one doing the analysis was present. I'll use aircraft accidents as the basis of my argument, but this argument is applicable to all domains of inquiry.
This is why commercial firms buy errors and omissions insurance. To protect the firm, their workers, and other professionals from claims of inadequate work or negligent actions. This is common in construction, consulting. Lloyds of London is a well known provider for this type of liability protection.
There are some who define Retrospective Coherence as a scientific process of interpreting the past.
Those making that claim have to redefine the term scientific process, since there is no underlying theory for assessing the observations or experiments to test against.
There are others who define it as sensemaking.
There are others who define it as pure fabrication to sell consulting services to business executives unaware of core management practices in the presence of uncertainties - both reducible and irreducible
One definition of Retrospective Coherence is it's an attempt to provide a scientific approach to understanding previous events. With it, practitioners within the “knowledge community” are formalizing concepts that historically fit under the age-old concept of “20 / 20 hindsight”.
The problem with looking backward is we always see the past through the biases of now. This is called cognitive bias.
And when doing this, we're laying the groundwork for failure to develop the correct understanding of the past and make informed decisions in the presence of uncertainty needed to increase the probability of success for the future. This approach is post facto and of little use once the failure has occurred.
Using Retrospective Coherence is an excuse for not doing the Root Cause Analysis and the Pre-Mortem Analysis needed to make informed decisions. Retrospective Coherence is an after the fact analysis of the crash. This does little for those who died in the crash. The approach to aircraft flight safety is to prevent accidents. It doesn't take much analysis to do that.
We have a phrase in the military flight community
The Pilot is the First to Arrive at the Scene of the Accident
Along with this, we had another phrase
Helicopters are very safe, just keep them away from the ground
This pre-mortem and root cause analysis (these are two sides of the same coin of Risk Management) is a proven approach to increasing the probability of success. Certain things are going to happen, certain things are going to happen. This means you need to have a plan when things do happen. This planning in depth is the basis of all high reliability, high integrity, fault tolerant, fail safe systems. Both physical, social, and soft systems. To suggest we only can deal with complex systems after the fact willfully ignores the principles and practices of Systems Engineering
Start with A Complexity Primer for Systems Engineers to learn how
The conjectures of soft solutions and fancy worded non-actionable outcomes proponents are post-facto providers. The airplane crash or the business crashed - where were they in the piloting of the aircraft or business? Answer - standing on the sidelines selling consulting courses
I work in a complex software-intensive system of systems domain, which includes flight systems, launch vehicles, power generation systems (nuclear and conventional), . On the programs we work, everyone is given a book to read - Disasters and Accidents in Manned Spaceflight. Along with Normal Accidents: Living with High Risk Technologies, The Next Catastrophe: Reducing our Vulnerabilities to Natural, Industrial, and Terrorist Disasters , and Failure is Not an Option.
So when we hear about retrospective coherence, the first question to ask is - How Could What We Are Observing with the 20 /20- hindsight have been foreseen if we had asked What Could Possibly Go Wrong?
The answer to that question starts with Root Cause Analysis and it's companion to Pre-Mortem Analysis. The first step in exploring this topic is to separate Social Systems from Engineered Systems. Yes Engineered Systems have people involved in their creation and use. But those interactions are dramatically different from the social systems of politics, societal interactions, and organizational complexity. There are lots of philosophical writings on this topic. But applying those to Increasing the Probability of Program and Project success has very little actionable process.
A Final Thought
When you hear the talk about retrospective coherence and it's biases and failures, search for the root cause of those failures. One root cause is poor leadership. I'm not talking about the touchy feely type of leadership we encounter in our modern software development. I'm talking about the leadership found here on the books in my office that we put to work every day:
- Leadership Strategy and Tactics
- Leaders Eat Last
- Modeling Complex Projects
- Making the Impossible Possible - Leading Extraordinary Performance: The Rocky Flats Story, where I was a program manager
- The Dichotomy of Leadership
- Extreme Ownership
- Red Team: How to Succeed by Thinking Like the Enemy
Enough of the theory of success from those who are self-proclaimed professors who are not practitioners bending metal into money.
Seek instead advice and guidance from those accountable for spending other people's where Failure is Not an Opinion.
